public class ContextFeatureExtractor extends FeatureCollection
To use this class, after creating the object, add other feature extractors using the
FeatureCollection.addFeatureExtractor(FeatureExtractor)
method. Then, for each neighboring word, the feature extractor will generate features using all
the extractors that have been added.
Note: This class only generates features from neighboring words. To go beyond words, new code needs to be written as required.
generators| Constructor and Description |
|---|
ContextFeatureExtractor(int contextSize,
boolean specifyIndex,
boolean ignoreConstituent)
Create a new ContextFeatureExtractor.
|
ContextFeatureExtractor(int contextSize,
boolean specifyIndex,
boolean ignoreConstituent,
WordFeatureExtractor... fex) |
| Modifier and Type | Method and Description |
|---|---|
Set<Feature> |
getFeatures(Constituent c) |
String |
getName() |
addFeatureExtractorpublic ContextFeatureExtractor(int contextSize,
boolean specifyIndex,
boolean ignoreConstituent)
contextSize - The number of tokens to the left and right of the constituent from which
the features should be extracted.specifyIndex - Should the feature mention the index (relative to the constituent)ignoreConstituent - Should the tokens in the constituent itself be ignored while
generating the features.public ContextFeatureExtractor(int contextSize,
boolean specifyIndex,
boolean ignoreConstituent,
WordFeatureExtractor... fex)
public Set<Feature> getFeatures(Constituent c) throws EdisonException
getFeatures in interface FeatureExtractorgetFeatures in class FeatureCollectionEdisonExceptionpublic String getName()
getName in interface FeatureExtractorgetName in class FeatureCollectionCopyright © 2017. All rights reserved.